robot speed

[cloud_254]

hey i'm tring to just get a ball park range for a robot's opitmum speed. any suggestion? o and btw wheres the white papers???

thx

[DonRotolo]

Hi! White papers are located in CD Media - look at the very top menu bar, just above the "Calendar" link.

As for robot speed, this very strongly depends on what the game is, how you want to play it, and what you expect to accomplish. As a point of reference, 1676's robot last year was designed for 11 feet per second. Our goal was to be stronger in defense (pushing ability) and not as stong in scoring ability (speed).

Don

[lukevanoort]

These are the are the ranges as I see it:
<4ft/s: Really, really slow
4-6ft/s: Medium slow, good torque, very good controllability
6-9ft/s: Medium speed, torque, controllability
9-10ft/s: Medium fast, decent torque, decent controllability
10-12ft/s: Fast, somewhat low torque, controllability depends on the drive design and driver
12-14ft/s: Very fast, low toque, probably needs a good driver
14-16ft/s: Extremely fast, very little torque, edge of controllability, needs a skilled driver
16+ft/s: Usually too fast. Very few teams have ventured into this territory except as the high gear of a shifter (254/968 come to mind) and I can't think of a successful single speed robot going this fast. Requires a quite good driver.

Of course, speed should be determined on a case by case basis; for example, in 2006 if your robot design was like 25's you could probably get by going 16ft/s even with a below average driver. This is because the turreted shooter only required that you get near the goal and you didn't have to maneuver finely to acquire balls. On the flip side, 2005's game required a lot of fine positioning so a low speed would probably be better. Also playing into the equation is your amount of traction and number of motors, but I wrote about that here so I'm not going to write it again.

[Cory]

Technically the optimum speed would be the combination of the fastest speed in which you can slip your wheels while pushing against a wall, and the speed at which you draw the least current.

For all practical purposes, without a lot of practice, or just plain gifted drivers, anything over 10-12 is too fast for most teams.

[alex1699]

i think fast but are team is a bunch of red necks

[AdamHeard]

If you had to pick a speed right now knowing nothing about the game; 6-10 fps while being still traction limited (being able to slip wheels like cory said) would be in the ballpark as universal. But, the much better choice is to wait to see what the game is, what your strategy requires and what type of drivetrain you have. An omni drive at 4 fps would be useless for most games, and a single speed 4wd at 14 fps would probably be bad as well for most games

Originally Posted by alex1699

i think fast but are team is a bunch of red necks

What????????????

EDIT: Don... I know he means our.... but his point is still pretty ambiguous.

[DonRotolo]

Originally Posted by alex1699

i think fast but are team is a bunch of red necks

I think he means "OUR team"....

[Richard Wallace]

Since alex1699's school in located in the middle of Connecticut, I doubt that his team is actually a bunch of rednecks. More likely it is a bunch of yankees.

I, on the other hand, attended a public high school in the middle of South Carolina. My school was quite literally full of rednecks. Some (most?) of my best friends were rednecks. (And the rest were transplanted yankees!) Rednecks are known for many quirks that non-rednecks might find strange, but a preference for slow gear ratios is emphatically NOT one of them.

Conclusion: Adam is right, the point is ambiguous at best.

[IndySam]

You might be a robot redkneck if:


You think subdivision is part of a math problem.

You take a load of scrap aluminum to the dump and bring back more than you took

Lego League was the best six years of your life.

Three quarters of the clothes you own have LOGOS on them.

[Gdeaver]

I'm a little surprised that all the posts mentioned FPS. A measure of velocity and no body had an comments on acceleration. What good is a robot that has a top speed of XX fps if it takes the entire length of the playing field to reach that speed. A robot with a low top speed and very quick acceleration can be very useful in some games.

[AdamHeard]

Originally Posted by Gdeaver

I'm a little surprised that all the posts mentioned FPS. A measure of velocity and no body had an comments on acceleration. What good is a robot that has a top speed of XX fps if it takes the entire length of the playing field to reach that speed. A robot with a low top speed and very quick acceleration can be very useful in some games.

this may be that it's kind of assumed that the robot is responsive and quick to accelerate; Well maybe it isn't, but I hope people would assume that . Definitely a good idea of you to point that out there, you never know what some people will do/assume....

I think in most cases, unless your going with some really high speeds, you will have very fast acceleration. Our 'bot this year had a top speed of 9.5 fps, and hit 9 fps in .2 seconds according to JVN's calc. I know in person it was slightly slower, but still no more than .5s to near top speed. Also, the in 05 and 06 we ran the old KOP trans with two small CIMs and had similar similar accelerations.

I think it'd be pretty hard to build a bot within a reasonable top speed that didn't accelerate quickly.

but then again, I've seen some strange things.........

[Richard McClellan]

Last year our bot was somewhat "acceleration limited" because if we went full speed forward, followed by full speed backward, we would tip over for sure because our arm extended out so far and was rather heavy on the end. We ended up limiting acceleration in the code to prevent some of our less experienced drivers from tipping over completely.

So, next year, we definitely hope to be able to design a robot with a much lower center of gravity so we don't have that problem anymore. But that is one reason that a robot could be "acceleration limited."

[kramarczyk]

Originally Posted by Gdeaver

I'm a little surprised that all the posts mentioned FPS. A measure of velocity and no body had an comments on acceleration. What good is a robot that has a top speed of XX fps if it takes the entire length of the playing field to reach that speed. A robot with a low top speed and very quick acceleration can be very useful in some games.

Originally Posted by AdamHeard

I think in most cases, unless your going with some really high speeds, you will have very fast acceleration. Our 'bot this year had a top speed of 9.5 fps, and hit 9 fps in .2 seconds according to JVN's calc. I know in person it was slightly slower, but still no more than .5s to near top speed. Also, the in 05 and 06 we ran the old KOP trans with two small CIMs and had similar similar accelerations.

We actually tune our bot's high gear to time over a distance as that is how most races are measured... i.e. 50 yard dash. Within JVN's drivetrain calculator (http://www.chiefdelphi.com/media/papers/1469) there is a modeler that allows you to figure how long it will take your bot to reach a reference distance. By iterating with you gear ratios you can determine a minimum time to a key distance in the game.

Example:
We had a 6x6 skid steer chassis with 8" wheels. Power was delivered via 2 small CIM's per side through AM gen 2 transmissions. This leaves us with only the final stage to 'futz' with in tuning the system. After looking at the game rules we decided that the key distance for the game was 13 ft. The packaging constraints of our system left us with a doman of sprockets that fit. Through the miracles of spreadsheets we iterate through each of the possible combinations and get a time to reach our reference distance. Then we just pick the sprockets that coorespond to the lowest time and call that our optimized high gear ratio. I have included a brief spreadsheet (http://www.chiefdelphi.com/media/papers/2046) that hopefully clarifies some of the grey spots in my description. In the spreadsheet I have also included a plot that shows how quickly the time increases as you move away from the optimum.

The low gear calcs are explained well in Ken and Paul's FRC Conference Presentation so I won't reiterate them here. They also address drivetrain performance with current limits. i.e. breakers Great stuff. http://www.chiefdelphi.com/media/papers/1682

Disclaimers:
- Yes, I realize that this optimized high gear ratio may need to be shifted to accomodate the low gear needs. This is an engineering decision based upon what your team deems important for the game.
- Yes, the iteration is annoying, but it takes less that 10 min to accomplish and the return on investment seems worth it. I have tried both the Goal Seek and Solver functions to automate this, but the table lookup that is used seems to cause them difficulty.

[Daniel_LaFleur]

Originally Posted by cloud_254

hey i'm tring to just get a ball park range for a robot's opitmum speed. any suggestion? o and btw wheres the white papers???

thx

This depends on:

1> what the challange is.
2> what strategy you choose.
3> The responsiveness and type of drive train you chose.
4> Skill of your drivers.
etc, etc, etc.

Last year, with us choosing to play defensive and all the scoreing done in the center of the field, so we chose high torque, high traction, and ~6.5 FPS.

[cloud_254]

good point but, i'm just trying to guess for this year's competition. i know kick off isn't til jan 5th but i'm just trying to get a head start.